full report - Deeresearch

advertisement
Final Client Report
Prepared for DEEResearch Ltd Contract No: 4.07
June 2005
“Wapiti Score” : A visual
assessment tool for determining
Wapiti introgression in herds of
weaner red deer.
J.F. Ward, I.C. Scott G.W. Asher, R.P. Littlejohn
AgResearch, Invermay Agricultural Centre,
Private Bag 50034, Mosgiel, New Zealand.
1
Client Report
This report has been prepared for DEEResearch Ltd and is CONFIDENTIAL to that
organisation and AgResearch. AgResearch will not disclose its contents to third parties unless
directed to do so by DEEResearch Ltd.
Every effort has been made to ensure this publication is accurate. However, because research
and development can involve extrapolation and interpretation of uncertain data, AgResearch will
not be responsible for any error or omission in this publication unless specifically agreed
otherwise in writing. To the extent permissible by law, neither AgResearch nor any person
involved in this publication accepts any liability for any loss or damage whatsoever that may
directly or indirectly result from any advice, opinion, representation, statement or omission,
whether negligent or otherwise, contained in this publication.
Dr Geoff Asher
Scientist – Integrated Agricultural Systems
Inquiries or requests to:
Jamie Ward
jamie.ward@agresearch.co.nz
Integrated Agricultural Systems Section
AgResearch – Invermay Agricultural Centre
Private Bag 50034, Mosgiel, New Zealand
2
Contents Page
Lay Summary ................................................................................................................. 4
Introduction .................................................................................................................... 5
Methods and Materials................................................................................................... 6
Animals and management ......................................................................................... 6
Measurements............................................................................................................ 6
DNA analysis .............................................................................................................. 7
Statistical analysis ...................................................................................................... 7
Results ........................................................................................................................... 7
Relationship between Wapiti Score and Elkmeter values ......................................... 7
Relationships of body measurements to Wapiti Score and Elkmeter values ............ 8
Accuracy of assigned Wapiti Scores.......................................................................... 9
Discussion .................................................................................................................... 10
Acknowledgements ...................................................................................................... 11
References ................................................................................................................... 12
Appendix 1: Wapiti scoring ......................................................................................... 13
Appendix 2: Wapiti scoring ......................................................................................... 15
3
Lay Summary

Poor reproductive rates from 2-year-old hinds, is a significant productivity issue
and reflects failure of many yearlings to attain puberty at 16 months of age.

Recent research (Asher et. al., 2004, 2005) has convincingly demonstrated that
the widespread introgression of Wapiti genes into the red deer herd is strongly
influencing pubertal performance through general failure of hinds of >20% wapiti
parentage attaining threshold liveweights for attainment of puberty.

This reflects a general failure to manage these animals as “Wapiti” (i.e. provide to
the specific needs of the genotype) in order to achieve their growth potential as
yearlings.

The present study develops and qualifies a “Wapiti Score” tool for farmers to
assess levels of Wapiti parentage in young (<1 year old) “red deer” to facilitate
genetic management decisions (e.g. early selection of red deer only replacement
females).

A 4-point scoring system is described that assigns levels of ”obviousness” of
Wapiti parentage to individual weaners (females and males analysed separately).
1 = no obvious Wapiti features
2 = slight obvious features
3 = distinct Wapiti features
4 = obvious Wapiti parentage

Scoring criteria outlined include a visual assessment of a pre-defined set of criteria
for stature, general pelage colouration, head and neck pelage characteristics and
rump patch size and colour (see Appendix 1 and 2).

Two independent observers each scored 449 rising-one-year-old “red deer” of
mixed genotype (202 females: 247 males) on three occasions; winter (early
August), spring (mid October) and summer (mid January) at 8-9, 10-11 and 13-14
months of age, respectively.

All weaners were also measured for shoulder height, body length and liveweight
on each occasion. A hair sample was collected from each weaner in August and
188 individuals (82 females: 106 males) were selected for DNA analysis of actual
Wapiti parentage using the GenometerTM G3 Hybrid Test (Elkmeter).

Both observers in the study exhibited highly significant positive correlations
between assigned Wapiti score and Elkmeter (DNA) values. The two observers
were generally in close agreement although one tended to be more conservative
in favour of red deer genotypes (i.e. slight underestimation of Wapiti parentage).

Both observers were more “accurate” in winter and spring (8-11 months of age)
than in summer (13-14 months). This probably reflects their visual assessment of
the seasonal pelage characteristics.

Wapiti Scores of 2-2.5 in winter and spring generally identified animals of ~20%
Wapiti parentage (the level considered to be biologically significantly different from
red deer from previous studies).

Morphometric measurements, either individual or in combination, were only
weakly correlated with Wapiti Score and Elkmeter, and did not provide additional
useful information.
4

Using data from two observers, we present measures of accuracy for “cut-off”
criteria of Wapiti Score to gauge the “risks” of assigning animal of the wrong
genotype.

“False positives” are the percentage of “red deer” that are incorrectly scored as
“Wapiti” (i.e. >20% Wapiti parentage).
”False negatives” are the percentage of ”Wapiti” that are incorrectly scored as “red
deer” (i.e. <20% Wapiti parentage).

Low score “cut-off” (i.e. WS=1) will generally ensure low incidence of retained
“Wapiti” but will likely see up to 20% of true “red deer” culled (i.e. false positives).
Conversely high score “cut-off” (i.e. WS 3) will exclude nearly all “red deer” but
see some true “Wapiti” culled (i.e. false negatives).
Conclusions

The “Wapiti Scoring” systems can be effective in identifying levels of Wapiti
introgression within herds of young “red deer” and for setting selection criteria for
the retention or culling of specific genotypes.

However, as with all subjective scoring systems, there are levels of inaccuracies
that must be recognised and accommodated.

While we provide guidelines for developing a scoring system, in reality each
observer will show individual experiential biases and impose different weightings
on the various trait characteristics. We recommend that those farmers serious
about applying “Wapiti Scoring” tools for long-term genetic management should
consider calibrating their personal scoring system using the GenometerTM G3
Hybrid Test.
Introduction
Within the last decade, the variable but generally poor reproductive performance of
young red deer hinds (rising-2-year olds) has been highlighted as a significant
productivity issue on New Zealand deer farms (Asher & Pearse, 2002).
Widespread use of ultrasound pregnancy scanning has revealed pregnancy rates of
young hinds ranging from 40-95% (Audige’ et. al., 1998; Beatson et. al., 2000). On the
poorer performing farms this is generally believed to reflect failure of many hinds to
reach puberty at 16 months of age (Asher & Pearse, 2002), despite the fact that average
liveweights at this age generally surpass the considered puberty liveweight threshold of
65kg for Western European genotypes (i.e. Cervus elaphus scoticus).
A recent study by Asher et. al,. (2004, 2005) has convincingly demonstrated that the
65kg threshold has little further relevance to the New Zealand deer industry due to the
widespread introgression of larger genotypes, particularly North American Wapiti (
Cervus elaphus roosevelti, nelsoni, manitobensis), that have higher puberty liveweight
thresholds. This study showed that under present red deer management systems; in
which average liveweights ranged between 87 and 100kg, yearling hinds with greater
than 20% wapiti genes generally failed to reach their genotype specific threshold for
attainment of puberty at 16 months of age, strongly biasing the non-pregnant population
at 18-20 months of age towards wapiti crossbred animals.
The key issue relating to improving reproductive performance of young hinds with >20%
wapiti genes is that of meeting the management requirements of these animals to
5
achieve their true growth potential, rather than having them compete with red deer
genotypes (<20% wapiti genes) in the same mob.
The study by Asher et. al., (2004, 2005) used “Wapiti Score”, as a subjective measure of
the “obviousness of wapiti characteristics” within individuals to assign variable levels of
introgression within populations of “red deer” hinds at 18-20 months of age. This 5 point
score (1 = red deer; 5 = very obvious wapiti crossbred) was assigned based on observer
assessment of general body conformation and pelage characteristics. It correlated
positively with more objective measures of introgression including the GenomnzTM DNA
based GenometerTM G3 Hybrid Test and morphometric data of body length, shoulder
height and liveweight. As such it proved to be a useful “ready reckoner” of wapiti
introgression, with scores of 3+ generally being applied to animals with >20% wapiti
genes.
This scoring system potentially serves as a useful tool for deer farmers to make
informed decisions and apply appropriate management practices to improve yearling
hind performance (e.g. separation of higher-order wapiti crossbreds to improve their
growth performance). However it needs to be adapted and assessed for use in younger
animals (e.g. weaners) for earlier application.
The present study aimed to develop and qualify a simpler “Wapiti Scoring” system for
application to populations of weaner “red deer” (i.e. 6-10 months of age).
Methods and Materials
Animals and management
A total of 449 rising-one-year-old red deer of mixed genotype were studied on a
commercial deer breeding and finishing farm in Otago. They were all born in Nov/Dec
2003 and were subjected to normal farm management practices.
The group consisted of 202 females of which at least 41 were of known wapiti
parentage. The remainder of the group (n=247) were males that were predominantly
individuals with some degree of wapiti parentage. All analyses treated females and
males as separate cohorts. The individuals of putative wapiti parentage (n=288) were
selected as a “bottom cut” by weight (mean 63kg); to ensure they would remain in the
trial before reaching killable weight (~100kg).
This also made them a more
homogeneous group than would normally be expected on-farm when performing the
Wapiti Scoring.
Measurements
All animals in the group were yarded for scoring and measuring on three occasions:

Winter (early August 2004) at approximately 8-9 months of age while in full winter
pelage.

Spring (mid-October 2004) at 10-11 months of age during pelage moult.

Summer (mid-January 2005) at 13-14 months of age while in summer pelage.
The animals were yarded off pasture for no more than 4 hours prior to measurements.
They were drafted in to a pen of 5-6 animals just prior to individual scoring and
measurements. All animals were then assessed for:
6
1.
‘Wapiti Score’; which was a 4-point ranking system looking for physical “wapiti”
features from a pre-defined set criteria (Appendix 1 & 2) for; stature, general
pelage colouration, head and neck pelage characteristics and rump patch size and
colour, with scores of; 1 = no obvious wapiti features, 2 = slight obvious wapiti
features, 3 = distinct wapiti features and 4 = obvious in their wapiti parentage.
Scoring was done independently by the same two observers throughout the trial.
2.
Shoulder height and body length as in Asher et. al., (2004, 2005)
3.
Liveweight on an electronic scale/weigh crate to 0.5kg resolution
During the first measurement period all animals had hair follicle samples collected prior
to measuring and scoring for later DNA testing. These samples were stored at ambient
temperature in paper envelopes until DNA testing in January 2005.
DNA analysis
A total of 82 females (including all 41 females from known wapiti parentage) and 106
males were DNA tested at the GenomnzTM DNA testing laboratory as described in
Asher et. al., (2004, 2005) (pp12-13) and analysed for their estimated percentage of
wapiti genes using the GenometerTM G3 Hybrid Test (Elkmeter).
Statistical analysis
Elkmeter score was regressed on individual and average observer Wapiti Score as both
a categorical and numerical variable. This was done separately for males and females
in winter, spring and summer.
Using a threshold Elkmeter score of 0.20 (20% Wapiti genes). A calculation was made
of the percentage of true “red deer” scored as Wapiti (i.e. false positives), and the
percentage of “Wapiti” scored as red deer (i.e. false negatives). This was done for each
observer.
Elkmeter value was also regressed on each morphometric variable: body length,
shoulder height, weight, shape (length/height), and Body Mass index (length x
height/weight). Again, this was done separately for males and females in winter spring
and summer.
In all cases, statistical significance was assessed at the 5% level. However, for the sake
of clarity, test results (significance levels) are not included in the Results.
Results
Relationship between Wapiti Score and Elkmeter values
Both observers in this study exhibited highly significant positive correlations between
assigned Wapiti Score (WS) and Elkmeter values. Generally, the two observers were in
close agreement, although Observer 2 tended to be more conservative, scoring more in
favour of red deer genotypes (i.e. lower WS). Table 1 provides the mean values for “%
Wapiti genes” against combined (average) WS of the two observers, as calculated from
the regression analysis.
For female weaners, the best correlations between WS and true genotype occurred in
winter and spring. The correlations in summer were considerably poorer but still highly
significant (Table 1). For male weaners, the spring correlations were slightly better than
in winter, but as for females, both were an improvement over summer. For both sexes,
7
mean WS in summer were generally lower, with more animals being scored as “red
deer” (i.e. WS=1) than in the previous periods.
Table 1: Mean “% wapiti genes”, calculated by regression analysis, for average Wapiti
Scores (WS) for the two observers.
(a)
Females
Mean
% Wapiti genes
Average WS1
Winter
Spring
Summer
1.0
8.9
8.8
8.9
1.5
11.8
11.1
13.4
2.0
13.8
15.2
18.0
2.5
23.4
31.0
32.7
3.0
39.7
34.3
30.5
3.5
40.3
50.0
-
4.0
-
-
-
6.0
6.0
8.0
RMS2
(b)
Males
Mean
% Wapiti genes
Average WS1
Winter
Spring
Summer
1.0
9.6
9.5
12.0
1.5
11.5
11.7
15.7
2.0
14.1
15.4
19.0
2.5
24.7
15.6
34.2
3.0
23.6
30.0
40.3
3.5
47.7
40.6
31.7
4.0
33.0
29.3
-
10.0
across the two observers
2 RMS = residual mean squared
10.0
10.0
RMS2
1 Average
Relationships of body measurements to Wapiti Score and Elkmeter values
Measurements of should height, body length and weight were analysed individually and
in combination to define “shape” (shoulder height / body length) and “body mass index”
(BMI; shoulder height x body length/weight). Generally, there were poor correlations
between such measurements and WS or Elkmeter. The best morphometric predictors
of wapiti parentage were body length for females in spring, and should height for males
in summer. However, in all cases WS alone was the best indicator of wapiti parentage,
and was not improved by adding morphometric variables.
8
Accuracy of assigned Wapiti Scores
Accuracy of WS is a measure of error likely to occur when selecting animals at or
beyond a given score for future herd replacements. In the case of a “red deer” only
replacement policy (e.g. removing wapiti influence from the future red deer hind
breeding herd), what are the risks of including animals of >20% wapiti genes at cut off
points of WS=1, 2 or 3?
Using data from the two observers in the study, estimates of sensitivity and specificity
are calculated (Table 2).
Table 2: Estimates of sensitivity and specificity of “cut=off” wapiti scores in terms of
false positives and false negatives for the two observers in the study,
False positives1
Wapiti Score
Obs. 1
Obs. 2
False negatives2
Obs. 1
Obs. 2
(a) Females
1
19.7%
19.7%
0%
9.1%
2
4.2%
1.4%
27.3%
36.4%
3
0%
0%
90.9%
81.8%
1
23.9%
22.5%
0%
9.1%
2
2.8%
1.4%
27.3%
45.5%
3
0%
0%
90.9%
81.8%
1
26.8%
22.5%
0%
27.3%
2
1.4%
2.8%
36.4%
72.7%
3
0%
0%
100%
100%
1
44.6%
54.2%
8.7%
8.7%
2
12.0%
6.0%
26.1%
52.2%
3
1.2%
0%
82.6%
87.0%
1
42.2%
44.6%
8.7%
0%
2
14.5%
8.4%
30.4%
21.7%
3
1.2%
1.2%
69.6%
82.6%
1
18.1%
12.0%
43.5%
39.1%
2
2.4%
2.4%
56.5%
73.9%
1.2%
0%
95.7%
3
1 % of “red deer” that are incorrectly scored as “wapiti”
2 % of “wapiti” that are incorrectly scored as “red deer”
100%
Winter
Spring
Summer
(b) Males
Winter
Spring
Summer
In Table 2; at a cut-off value of WS=1, it can reasonably be expected that only between
0 and 10% (depending on observer) of retained females in winter will be of “wapiti”
genotype (i.e. false negatives). However, nearly 20% of true “red deer” genotype will be
called as “false positives” (i.e. erroneously scored as “wapiti” type). As the “cut-off”
9
value increases (i.e. WS 2), more “red deer” are retained; but more “wapiti” are also
included in the selected group.
This process can also be performed in reverse in order to select animals with >20%
wapiti parentage. Clearly, selection of animals at or above WS=3 will include the vast
majority of “wapiti” genotype while excluding nearly all “red deer” genotype (Table 2).
However, selection at and above WS=2 will increase “risks” of false negatives.
Discussion
The value of any subjective scoring system lies principally in ease of field application
and low cost. In this respect, the ”Wapiti Score” system outlined in this report meets
these criteria.
However, it must be recognised that Wapiti Score is an assessment of genotype based
on visual appraisal of phenotype. As such, it is subject to a degree of inaccuracy
because animal phenotype does not always reflect true genotype. Environmental
effects and “random segregation of alleles” will influence expression of genotype. There
will always be a “grey area” where animals are wrongly assigned to genotype (we refer
to this in this report as “false positives” and “false negatives”).
Fortunately, red deer and Wapiti subspecies are morphologically very distinct
genotypes. They exhibit extreme differences in stature, size, shape and, quite
importantly, pelage colour and appearance. This means that in the process of
hybridisation (more correctly termed “crossbreeding”) a variety of morphological markers
of genotype are expressed in various combinations by individuals. Even at quite low
levels of Wapiti parentage (e.g. <20%) certain “Wapiti” characteristics prevail in the
phenotype.
The “Wapiti Score” essentially pieces together a visual assessment of a number of
“Wapiti” traits that scores animals on the “obviousness of wapiti-ness” (Asher et. al.,
2004, 2005). Each trait alone is probably a poor indicator of “wapiti-ness” but in
combination provide a fairly accurate assessment of introgression of the Wapiti
genotype. Perhaps the most useful traits are those of pelage characteristics (e.g.
colour) as, unlike body size parameters, these are less subject to environmental
influence. However, due to the probability that relatively few genes control pelage
colour, the process of “random segregation of alleles” may eliminate some wapiti pelage
characteristics from a few individuals with a high level of Wapiti parentage. This simply
highlights the importance of not relying on only one aspect of physical appearance when
scoring animals.
The results of the present study show that when a number of “wapiti-like” physical traits
are combined into a scoring system, we can obtain a reasonably good assessment of
the level of true Wapiti introgression within a herd and we can further use that to set
decision criteria for the retention or removal of individuals in order to meet specific
genetic management plans (e.g. selection of red deer only replacement females).
The 4-point scoring system presented in this report is a simplification of the 5-point
system used for yearling hinds by Asher et. al., (2004, 2005). It has been specifically
designed to use in younger deer (i.e. weaners) to facilitate earlier decision-making (i.e.
well before establishment of mating groups). While it is arguably most useful for
selecting female replacements, we also assessed its effectiveness in male weaners
should farmers wish to separate out discrete genetic lines for breeding or more efficient
venison production systems
The results of the present study demonstrate that the “Wapiti Scoring” system is best
applied during winter (6-8 months) and spring (9-11 months) rather than summer (12-14
10
months). This reflects the “accuracy” as defined by the regression of Elkmeter
(objective measure of Wapiti parentage) on Wapiti Score (subjective measure).
Observers in the present study tended to under-estimate the level of Wapiti influence
within individuals in summer (i.e. more animals had WS=1 than on previous occasions),
particularly for males. This in turn, probably reflects the assessment of the pelage traits,
with early summer pelage colour and texture being less differentiated between red deer
and Wapiti.
Interestingly, the morphometric data, (shoulder height, body length and liveweight), for
the young deer in the present study did not provide any further useful information on
Wapiti introgression.
Correlations between individual/combined parameters and
Elkmeter were weak (i.e. generally not statistically significant). This is in marked
contrast to earlier studies on yearling (18-20 month old) hinds, in which shoulder height
and body length were useful additional predictors of “wapiti-ness” (Asher et. al., 2004,
2005). This is probably due to a couple of facts: most importantly that the weaners in
the present study were actively growing throughout the study period, with large
population variances on individual growth performance at any given point in time.
Yearling red deer hinds in the previous study had probably achieved 85-90% of their
ultimate body size by the age when they were measured, with body size/shape better
reflecting their true genotype. Secondly that the selection of the group by weight
probably reduced or skewed the expected normal distribution of sizes and weights. This
highlights the importance of assessing a wide range of physical characteristics when
assigning Wapiti Scores to the younger cohorts.
In conclusion, while the 4-point Wapiti Scoring System is potentially a useful tool that will
allow farmers to make reasonably objective decisions on genetic management at an
early stage (i.e. weaners), cognisance must be given to the inherent inaccuracies of
subjective scoring of phenotype. For any given cut-off score there will be a risk that
some inappropriately assigned animals will be either retained or culled. Clearly, the
degree to which this will happen will depend on, firstly, the “cut-off” level decision and,
secondly, the ability of the observer to perform the scoring. The latter issue is very
important because, at the end of the day, each observer will make personal judgements
on what trait characteristics define “wapiti-ness”. Each individual, through personal
experiential bias, will impose different weightings on the various trait characteristics.
While we provide a detailed guide on how to apply the 4-point scoring system (Appendix
1 and 2), each observer will still need to develop their own personal system based on
experience of the ‘type’ of their animals.
We would recommend that farmers who are serious about applying the Wapiti Scoring
System for long-term genetic management of their herd should consider, at some stage,
calibrating their personal scoring system using the Elkmeter test. This would entail DNA
(hair) sampling and analysis of a selection of individuals of each Wapiti score, and
correlating the two measures. This would serve to define biases and make more
informed decisions on cut-off criteria.
Acknowledgements
We thank Landcorp; especially Steve Mitchell and staff and Landcorp’s Hindon Station
for providing access to the weaners, use of their yards and labour in the yards during
this study and the Foundation for Research Science and Technology for co-funding this
project.
11
References
Asher, G.W. and Pearse, A.J. (2002) Managing reproductive performance of farmed
deer: the key to productivity.
Proceedings of the Third World Deer Farming
Congress, Austin, Texas, USA. 99-112.
Asher, G.W., Archer J.A., Scott I.C., O’Neill K.T., Ward J. and Littlejohn R.P. (2004)
Reproductive performance of pubertal red deer (Cervus elaphus) hinds: Effects of
genetic introgression of wapiti subspecies on pregnancy rates at 18 Months of
age. Report to DEEResearch
Asher, G.W., Archer J.A., Scott I.C., O’Neill K.T., Ward J. and Littlejohn R.P. (2005)
Reproductive performance of pubertal red deer (Cervus elaphus) hinds: Effects of
genetic introgression of wapiti subspecies on pregnancy rates at 18 Months of
age. Animal Reproduction Science (in press).
Audigé, L.M., Wilson, P.R. and Morris, R.S. (1998) A body condition score system and
its use for farmed red deer hinds. New Zealand Journal of Agricultural Research
41:545-553.
Beatson, N.S., Campbell, A.C. and Judson, H.G. (2000) Deer Industry Manual New
Zealand, Herald Communications Ltd, Timaru, NZ: 134 pp.
12
Appendix 1: Wapiti scoring
Wapiti Scoring Red vs. Wapiti extremes weaners in winter
Red (WS = 1)
Wapiti (WS = 4)
Stature
Finer, shorter, more
proportionate body and limbs.
Tall, thicker blockier more rectangular
body. Legs appear more gangly,
awkward and less proportionate to
body.
Colour
Red and grey mostly.
Grey face and underside, red
body
Chocolate-black and cream.
Head and neck (throat mane) darker
body lighter to cream.
Head
More elongated appearing
muzzle, grey face, red to brown
chevron
Pointed ears
Muzzle more stout/shorter appearing,
brown face and chevron.
Rounded fuzzy ears.
Rump
Smaller, less apparent rump
patch, any colour from fully red
to fully white often red top and
white bottom.
Tail different colour.
Patch shaped like an upside
down pear.
Large obvious creamy white rump
patch.
Tail same colour as rump.
Heart shaped – quite rounded
looking.
13
14
Appendix 2: Wapiti scoring
Wapiti Scoring red vs. Wapiti extremes in summer (yearling hinds)
Red (WS = 1)
Wapiti (WS = 4)
Stature
Finer, shorter, more
proportionate body and limbs
Tall, thicker blockier more rectangular
body, deeper chest. Legs appear
longer and less proportionate to body
Colour
Red and grey mostly.
Red flecked chevron, grey face,
muzzle, neck front and
underside, red flecked body.
Colour lightens from back to
belly
Rusty red, brown and cream.
Head red and brown, neck (throat
mane) brown; darker than body.
Body quite uniform rust red,
underside mostly cream.
Head
Grey face, red to brown
chevron.
Longer skinnier ears.
Rusty red-brown face and chevron,
nose and muzzle dark brown.
Shorter wider ears
Rump
Smaller, less apparent rump
patch, any colour from fully red
to fully white often red top and
white bottom.
Tail different colour.
Less obvious viewed from side
on.
Larger creamy white rump patch.
Tail same colour as rump.
Obvious viewed from side on.
15
16
Download